A Comparative Study of Brain-Computer Interface Paradigms in Lower Limb Motor Rehabilitation

Abstract

Lower limb motor dysfunction resulting from neurological injuries presents a significant clinical and engineering challenge. Brain-Computer Interface (BCI) technology offers a direct neural pathway for controlling assistive devices, yet the comparative efficacy of different BCI paradigms remains insufficiently quantified. This study presents a systematic engineering evaluation of three primary BCI paradigms-P300, Steady-State Visual Evoked Potential (SSVEP), and Motor Imagery (MI)-applied to lower limb rehabilitation. We analyze their performance across four quantitative dimensions: walking ability, physiological function, motor control, and quality of life. Clinical data analysis reveals that MI-based systems combined with physical training yield the most significant improvements in muscle strength (e.g., hip flexor strength increased from 2.58±0.44 kg to 3.46±0.66 kg over 4 weeks, p<0.001). P300 paradigms demonstrate high stability for long-term function maintenance, evidenced by significant amplitude increases (from $6.16 \pm 3.34 \mu \mathrm{V}$ to $9.52 \pm 2.66 \mu \mathrm{V}, \mathrm{p}=0.001$) correlating with neural recovery. SSVEP systems excel in high-precision gait training due to their robust frequency response. Furthermore, hybrid paradigms (e.g., MI-SSVEP) show superior potential for enhancing neural plasticity. This comparative analysis provides a technical framework for selecting and optimizing BCI paradigms based on specific rehabilitation engineering requirements.